| What can AI teach me about training people? | The way you train an AI for your organization is also the way you should train people: give them context, time, immediate feedback, and break work into manageable chunks. |
The AI industry is spending billions of dollars to train artificial intelligence models, betting big that it can profit from tools that might aid (or replace) human efforts.
To give you a sense of the scale of this effort: By the end of 2025, big tech will have spent more on AI development than the combined budgets for three of the most expensive research projects in human history (the Manhattan Project, the International Space Station, and the Space Shuttle program).
Meanwhile, in nonprofit and other public service organizations, training does not get the same love.
Rather than being seen as a necessary investment in the future, training is often seen as pure cost: cash spent on courses, time spent away from core work, etc. Funders and advisors even treat training dollars as “overhead and administration” expenses, costs that take away from the mission.
All AI hype aside, the difference in the old and new attitudes to training begs a question: what would happen if we showed the same level of enthusiasm for training people as we’re showing for training AI?
Because training isn’t a cost; it’s a lifeline. Training and development lead to financial stability, to greater impact, or both.
As much as it pains me to agree with the AI bros, I’m with them on one thing: if we want our organizations and our sector to succeed, we need to invest strategically in training and development.
So let’s get excited.
AI productivity tools are not the magic solutions many expect them to be. Before they actually work for your organization, AI tools need to be trained and prompted and cajoled into “understanding” your unique needs.
The sales headline screams “Our tool will change your life!” The fine print whispers, “Once you spend some time doing X,Y, Z.”
As it turns out, “X, Y, and Z” are not small tasks. The setup to make an AI system go requires effort and intentionality … almost identical to what you’d need to do to train and support humans.
And that’s the interesting thing: the way you train an AI for your organization is the way you’d train people. In that sense, the most interesting contribution AI is making to organizational life right now is providing a blueprint for ways managers can be more effective.
Whether you use AI tools or not, the insights into training and development are super helpful.
Your team members need background information to understand what your organization is all about. Give them your mission and goals, explain which products and services you provide, who uses them, who funds them, how you communicate, and what sets you apart.
If you heard that off-the-shelf ChatGPT or Gemini would be able to magically skip this step, you’d be wrong. With machines, you still have to pull all the relevant information together, organize it, feed it to the model, and check to see if it’s been retained correctly.
Likewise, even the most experienced fundraiser or scientist who joins your team will need access to specific information that is unique to your organization. So yep, you still need those organizational history documents, onboarding packages, policies, procedures, and style guides.
Background information alone won’t set your team up for success on their tasks and duties. Give them specific guidance and detail so that they can deliver work that is effective and fits with your organization’s approach: Do you have last year’s critical path for a project? A recent program evaluation? Detailed notes in your CRM for that upcoming donor meeting? Benchmarks for online engagement? Share it!
AIs work exactly the same way. Unless you provide sufficient context for your request, the work you get from an AI will be off-target, mediocre, or a carbon copy of someone else’s work. Skills like “prompt engineering” and “context engineering” have less to do with AI and more to do with setting up others for success — humans included.
Once you’ve shared all this context with someone, they need time to review the information and internalize it. Then they need time to complete the task that you’ve requested.
With an AI, it seems obvious that you can’t interrupt it. The little spinny thing on the screen tells you “working …. thinking … doing stuff …”. If you enter more information into the box, it’ll stop doing the thing it was doing or ignore you entirely. For complex tasks, AI tools even say something like “This is going to take a while. Go do something else and come back to me later.”
Your human team members need the same space. Don’t interrupt them, toss out a different instruction, and expect them to complete the first task correctly. Don’t assign a priority task, then immediately ask for four other, unrelated things. Set clear priorities, then give your people time and space to do the work. Happily, this is also how you build trust, autonomy, and confidence.
The more quickly your team receives feedback, the more they’re able to learn from it. It’s tough with a busy nonprofit schedule, but your people need feedback when the work is still fresh. Timing matters.
When giving feedback, assess whether the work is good enough, or whether more refinement is necessary. Most projects are iterative, so it should feel comfortable and natural to make some changes after a first review.
If something needs more work or correction, be clear about what needs to be fixed and give your team time to go off and create the next version.
Again, note the pattern: the way you’d give feedback to your AI is exactly the same as what your people need from you as their manager!
Both people and technology can short-circuit when work becomes overwhelming. From short-tempered co-workers to computer program crashes, you’ve seen it happen. When overwhelm hits, the solution is to break the work down further. Simplify.
One of the main things you should do as a manager is figure out how to break work down into ideal chunks. You want tasks that are big enough to move a project forward, but focused enough to be done skillfully and sustainably.
With AI, you get immediate feedback when your work chunks are too complex: the results are terrible, the tool freezes, or your data consumption (and the associated bill) soars.
With your human team members, the feedback may be harder to interpret. You may worry the person isn’t “up to” the task or can’t understand the instructions. You may think they aren’t experienced enough yet and start to lower your expectations. Before you make those kinds of judgements, review your instructions and the context you provided. Try to clarify them, focus them, or reframe them and give your team member another shot.
Every path forward requires the same core management skills, whether you’re trying to lead a team or adopting AI. So as a starting point, it’s worth cultivating the skills that will help you manage effectively.
Meanwhile, it seems only fair to give people the same opportunities we’d give machines. It certainly can’t hurt.
The Organizer is a newsletter for people working to create equitable and sustainable communities. Whether you are part of a nonprofit, a charity, or a social enterprise, this newsletter is for you.
Each edition, we explore one aspect of social impact work. We answer a common “How do I …?” question, and we tell you about a tool that will help make your work a little easier. Subscribe for free at Entremission.com.